Top Banner
This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 6699 Cite this: Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 Mechanisms and advancement of antifading agents for fluorescence microscopy and single-molecule spectroscopywz Thorben Cordes,* ab Andreas Maiser, c Christian Steinhauer, a Lothar Schermelleh* c and Philip Tinnefeld* ad Received 23rd September 2010, Accepted 14th January 2011 DOI: 10.1039/c0cp01919d Modern fluorescence microscopy applications go along with increasing demands for the employed fluorescent dyes. In this work, we compared antifading formulae utilizing a recently developed reducing and oxidizing system (ROXS) with commercial antifading agents. To systematically test fluorophore performance in fluorescence imaging of biological samples, we carried out photobleaching experiments using fixed cells labeled with various commonly used organic dyes, such as Alexa 488, Alexa 594, Alexa 647, Cy3B, ATTO 550, and ATTO 647N. Quantitative evaluation of (i) photostability, (ii) brightness, and (iii) storage stability of fluorophores in samples mounted in different antifades (AFs) reveal optimal combinations of dyes and AFs. Based on these results we provide guidance on which AF should preferably be used with a specific dye. Finally, we studied the antifading mechanisms of the commercial AFs using single-molecule spectroscopy and reveal that these empirically selected AFs exhibit similar properties to ROXS AFs. 1. Introduction Fluorescence light microscopy has become an indispensable tool in various scientific fields, ranging from biomedical research to material sciences. One of its key features is the possibility to specifically label and detect structural components of interest with spectrally distinct fluorophores, e.g., to analyze the spatial distribution of biomolecules within cells and tissues by immunohistochemistry. 1 More recently, the introduction of single-molecule approaches 2,3 and super-resolution imaging techniques, 4–7 have further extended the capabilities and range of applications. Modern fluorescence applications are strongly dependent on the performance and characteristics of the fluorescent probes used. Important properties of the labels are their brightness (given by the product of extinction coefficient at the excitation wavelength and the fluorescence quantum yield), photostability (i.e., resistance to irreversible, light-induced reactions), storage stability of stained samples, solubility in water (for biological applications), the capability to chemically link the label to the structure of interest and finally a minimized influence of the label onto the labeled structure itself. Organic fluorophores often represent the premier choice due to their brightness, chemical flexibility, small size and many new labeling strategies even for in vivo applications. 1,8 Many substances, especially reductants such as ascorbic acid (AA), 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox, TX), 9,10 p-phenylenediamine (PPD, used in the commercial product Vectashield), 1,4-diazabicyclo[2.2.2]octane (DABCO, used in Ibidi-MM) 11 or n-propyl gallate and triplet-quenchers such as mercaptoethylamine 12 and cyclo- octatetraene, 12,13 have been known to improve the photo- stability of fluorophores for microscopy and single-molecule spectroscopy. 14,15 Antifading substances are typically dissolved in glycerol-buffer or aqueous solution, which preserve the sample morphology. It should be noted that other frequently used embedding media have polymerizing formulae (e.g., Prolong Gold, Moviol). While good antifading properties have been reported, 14,15 cells embedded in these hardening media show substantial flattening, which is often disadvantageous for applications where the preservation of the 3-dimensional (3D) morphology is a requirement. Herein, we compare a a Applied Physics – Biophysics & Center for NanoScience (CeNS), Ludwig Maximilian University of Munich, Amalienstr. 54, 80799 Munich, Germany. E-mail: [email protected], [email protected]; Fax: +49 531 391 5334 b Biological Physics Research Group, Department of Physics, University of Oxford, Clarendon Laboratory, Parks Road, Oxford OX1 3PU, United Kingdom c LMU Biocenter, Department of Biology, Ludwig Maximilian University of Munich, Grosshaderner Str. 2, 82152 Planegg-Martinsried, Germany. E-mail: [email protected] d NanoBioSciences, Institute of Physical and Theoretical Chemistry, TU Braunschweig, Hans-Sommer-Str. 10, 38106 Braunschweig, Germany w This article was submitted as part of a Themed Issue on single- molecule optical studies of soft and complex matter. Other papers on this topic can be found in issue 5 of vol. 13 (2011). This issue can be found from the PCCP homepage http://www.rsc.org/pccp. z Electronic supplementary information (ESI) available. See DOI: 10.1039/c0cp01919d PCCP Dynamic Article Links www.rsc.org/pccp PAPER Downloaded by Ludwig Maximilians Universitaet Muenchen on 25/04/2013 13:08:05. Published on 11 February 2011 on http://pubs.rsc.org | doi:10.1039/C0CP01919D View Article Online / Journal Homepage / Table of Contents for this issue
11

Citethis: Phys. Chem. Chem. Phys .,2011, PAPER€¦ · Citethis: Phys. Chem. Chem. Phys .,2011, ... aApplied Physics ... also find more specific effects that we summarize in practical

Jun 06, 2018

Download

Documents

truonglien
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Citethis: Phys. Chem. Chem. Phys .,2011, PAPER€¦ · Citethis: Phys. Chem. Chem. Phys .,2011, ... aApplied Physics ... also find more specific effects that we summarize in practical

This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 6699

Cite this: Phys. Chem. Chem. Phys., 2011, 13, 6699–6709

Mechanisms and advancement of antifading agents for fluorescence

microscopy and single-molecule spectroscopywzThorben Cordes,*

abAndreas Maiser,

cChristian Steinhauer,

aLothar Schermelleh*

c

and Philip Tinnefeld*ad

Received 23rd September 2010, Accepted 14th January 2011

DOI: 10.1039/c0cp01919d

Modern fluorescence microscopy applications go along with increasing demands for the employed

fluorescent dyes. In this work, we compared antifading formulae utilizing a recently developed

reducing and oxidizing system (ROXS) with commercial antifading agents. To systematically

test fluorophore performance in fluorescence imaging of biological samples, we carried out

photobleaching experiments using fixed cells labeled with various commonly used organic dyes,

such as Alexa 488, Alexa 594, Alexa 647, Cy3B, ATTO 550, and ATTO 647N. Quantitative

evaluation of (i) photostability, (ii) brightness, and (iii) storage stability of fluorophores in

samples mounted in different antifades (AFs) reveal optimal combinations of dyes and AFs.

Based on these results we provide guidance on which AF should preferably be used with a specific

dye. Finally, we studied the antifading mechanisms of the commercial AFs using single-molecule

spectroscopy and reveal that these empirically selected AFs exhibit similar properties to

ROXS AFs.

1. Introduction

Fluorescence light microscopy has become an indispensable

tool in various scientific fields, ranging from biomedical

research to material sciences. One of its key features is the

possibility to specifically label and detect structural components

of interest with spectrally distinct fluorophores, e.g., to analyze

the spatial distribution of biomolecules within cells and tissues

by immunohistochemistry.1 More recently, the introduction of

single-molecule approaches2,3 and super-resolution imaging

techniques,4–7 have further extended the capabilities and range

of applications. Modern fluorescence applications are strongly

dependent on the performance and characteristics of the

fluorescent probes used. Important properties of the labels

are their brightness (given by the product of extinction coefficient

at the excitation wavelength and the fluorescence quantum

yield), photostability (i.e., resistance to irreversible, light-induced

reactions), storage stability of stained samples, solubility in

water (for biological applications), the capability to chemically

link the label to the structure of interest and finally a minimized

influence of the label onto the labeled structure itself. Organic

fluorophores often represent the premier choice due to their

brightness, chemical flexibility, small size and many new

labeling strategies even for in vivo applications.1,8

Many substances, especially reductants such as ascorbic

acid (AA), 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic

acid (Trolox, TX),9,10 p-phenylenediamine (PPD, used in the

commercial product Vectashield), 1,4-diazabicyclo[2.2.2]octane

(DABCO, used in Ibidi-MM)11 or n-propyl gallate and

triplet-quenchers such as mercaptoethylamine12 and cyclo-

octatetraene,12,13 have been known to improve the photo-

stability of fluorophores for microscopy and single-molecule

spectroscopy.14,15 Antifading substances are typically dissolved

in glycerol-buffer or aqueous solution, which preserve the

sample morphology. It should be noted that other frequently

used embedding media have polymerizing formulae (e.g., Prolong

Gold, Moviol). While good antifading properties have been

reported,14,15 cells embedded in these hardening media show

substantial flattening, which is often disadvantageous for

applications where the preservation of the 3-dimensional

(3D) morphology is a requirement. Herein, we compare a

a Applied Physics – Biophysics & Center for NanoScience (CeNS),Ludwig Maximilian University of Munich, Amalienstr. 54,80799 Munich, Germany. E-mail: [email protected],[email protected]; Fax: +49 531 391 5334

b Biological Physics Research Group, Department of Physics,University of Oxford, Clarendon Laboratory, Parks Road,Oxford OX1 3PU, United Kingdom

cLMU Biocenter, Department of Biology, Ludwig MaximilianUniversity of Munich, Grosshaderner Str. 2,82152 Planegg-Martinsried, Germany.E-mail: [email protected]

dNanoBioSciences, Institute of Physical and Theoretical Chemistry,TU Braunschweig, Hans-Sommer-Str. 10, 38106 Braunschweig,Germany

w This article was submitted as part of a Themed Issue on single-molecule optical studies of soft and complex matter. Other papers onthis topic can be found in issue 5 of vol. 13 (2011). This issue can befound from the PCCP homepage http://www.rsc.org/pccp.z Electronic supplementary information (ESI) available. See DOI:10.1039/c0cp01919d

PCCP Dynamic Article Links

www.rsc.org/pccp PAPER

Dow

nloa

ded

by L

udw

ig M

axim

ilian

s U

nive

rsita

et M

uenc

hen

on 2

5/04

/201

3 13

:08:

05.

Publ

ishe

d on

11

Febr

uary

201

1 on

http

://pu

bs.r

sc.o

rg |

doi:1

0.10

39/C

0CP0

1919

D

View Article Online / Journal Homepage / Table of Contents for this issue

Page 2: Citethis: Phys. Chem. Chem. Phys .,2011, PAPER€¦ · Citethis: Phys. Chem. Chem. Phys .,2011, ... aApplied Physics ... also find more specific effects that we summarize in practical

6700 Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 This journal is c the Owner Societies 2011

recently developed antifading formula that simultaneously

employs reducing and oxidizing agents16 with common, com-

mercial, glycerol-based antifading formulae suited for (3D)

fluorescence microscopy.11,14,15

In contrast to the empirically found AFs, a recently established

formula using a reducing and oxidizing system (ROXS) is

based on a photophysical model. The underlying rationale

is that an antifading formula has to rapidly and effectively

depopulate reactive intermediate states other than the

singlet states S0 and S1 that are part of the fluorescence

excitation-emission cycle. Every time the fluorophore enters

a potentially reactive state of a lifetime significantly longer

than S1 it should be depopulated to ultimately restore the

ground state (see Fig. 1 for energy level diagram).

This should not only increase photostability but dependent

on the excitation conditions also increase the brightness of the

fluorophore. The ROXS concept realizes this rationale using

electron transfer reactions. As soon as a transient state of

higher energy—which is a potential reactive intermediate on

common photobleaching pathways such as the triplet state—is

formed, it is depopulated through an electron transfer reaction

either with the reductant or with the oxidant. In the case of a

reduction, a radical anion is formed whereas in the case of a

primary reaction with an oxidant the corresponding radical

cation is formed. In the following, these transient states are

also rapidly depopulated through the complementary electron

transfer reaction, that is, the radical anion will react with the

oxidant to form the ground state. Alternatively, the radical

cation reacts with the reductant also yielding the ground state.

In all instances, the ground state is rapidly formed and the

build-up of a broad range of transient states including triplet

and radical states is suppressed (see scheme in Fig. 1). These

reactions are commonly diffusion controlled.

The ROXS concept was demonstrated to efficiently eliminate

blinking (due to triplet states as well as redox blinking due to

radical ion states) from single-molecule intensity transients for

a variety of different fluorophore classes (cyanines, (carbo-)

rhodamines, oxazines) by different combinations of oxidants

and reductants.10,16–18 For most fluorophores, particularly in

the yellow and red spectral regions a drastic increase of

photostability was observed.18 Targeted induction of transient

dark states by adapting the concentrations of redox active

agents was also exploited for super-resolution imaging by

different approaches.19–21 But the implementation of ROXS

in common biological imaging, e.g., confocal microscopy,

Fig. 1 Jablonski diagram according to the ROXS concept: A single

fluorophore is cycled between its singlet ground state (S0) and the

fluorescent first excited state (S1) with the excitation rate kexc emitting

a characteristic number of photons. The probability to enter the dark

triplet state (T1) with its inverse lifetime kT is given by the rate constant

of intersystem crossing (kisc). As the high reactivity of the triplet state

can cause irreversible destruction of the fluorophore it is rapidly

depleted with rates kred0/ox by oxidizing or reducing agents forming a

radical cation or anion (F��), respectively. Each of the ionized states is

depleted by the complementary redox reaction with rate constants

kred0 /ox0 recovering the electronic ground state.

Scheme 1 Molecular structures of fluorophores investigated in the present study: ATTO 647N,22 Alexa 488/594,23 Cy3B,24 Alexa 647.25 The dyes

are sorted according to the class of fluorophores (carbo-rhodamines, rhodamines, cyanines). The structure of ATTO 550 was not available.

Dow

nloa

ded

by L

udw

ig M

axim

ilian

s U

nive

rsita

et M

uenc

hen

on 2

5/04

/201

3 13

:08:

05.

Publ

ishe

d on

11

Febr

uary

201

1 on

http

://pu

bs.r

sc.o

rg |

doi:1

0.10

39/C

0CP0

1919

D

View Article Online

Page 3: Citethis: Phys. Chem. Chem. Phys .,2011, PAPER€¦ · Citethis: Phys. Chem. Chem. Phys .,2011, ... aApplied Physics ... also find more specific effects that we summarize in practical

This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 6701

and a comparison with common AFs used for immunofluores-

cence is still missing.

In this paper, we compare the performance of two ROXS

AFs (ascorbic acid/methyl viologen, Trolox/Trolox-quinone)

with common substances also used in commercial AFs such

as phenylenediamine and DABCO. We therefore carried

out spinning disk confocal microscopy in combination with

repetitive point laser scanning for the imaging and controlled

bleaching of immunolabeled nuclear proteins in fixed mammalian

tissue culture cells. From these data sets we extracted the

parameters (i) photostability and (ii) brightness. By measuring

the same samples again after three days of storage we

additionally obtained the (iii) storage stability in the respective

mounting medium. To test for the general validity of our

results we characterized various organic dyes of different dye

classes, such as cyanines and (carbo)-rhodamines, spanning

the entire visible spectrum. For all studied dyes (Alexa 488,

Alexa 594, Alexa 647, Cy3B, ATTO 550, ATTO 647N),

available molecular structures are shown in Scheme 1.

Our data suggest that the ROXS based AFs significantly

improve the performance of dyes in immunofluorescence

imaging and often outperform classical AFs. We, however,

also find more specific effects that we summarize in practical

guidance outlining which particular AF should be used for a

given fluorophore (class).

To learn more about the mechanism of the empirically

found antifading agents we finally used single-molecule

spectroscopy and studied the influence of different AF con-

centrations on the blinking of single dye molecules. Therefore,

we used the electron affine oxazine ATTO 655 that has

been termed single-molecule redox sensor (SMRS) because it

selectively reports on the oxidizing and reducing properties of

the environment: the lifetime of the dark state reports on the

oxidizing properties of the medium whereas the number of

emitted photons for one ON-state reports on the reducing

properties of the medium.10 Shortening OFF-times at con-

stant emitted photons of the SMRS have recently revealed

the antifading mechanism of Trolox (TX) that forms a

quinone over time and then acts according to the ROXS

concept.10

For all commercial AFs we found shortening OFF-state

lifetimes at increasing AF-concentration indicating unexpected

and substantial oxidizing capabilities of the AFs that are

considered to be reductants. This leads to the conclusion that

their empirical selection might be related to their additional

oxidizing properties and that their stabilizing mechanism is

also related to the ROXS concept. This fact of course under-

lines the importance of the ROXS concept not only for single-

molecule spectroscopy but also for fluorescence applications in

general.

2. Material and methods

Immunostaining, confocal microscopy and photobleaching

experiments

Human HeLa cells stably expressing histone H2B-GFP, and

HeLa Kyoto cells were cultured in DMEM supplemented

with 10% fetal bovine serum and 50 mg ml�1 gentamycine.

For photostability experiments cells were grown to 50–80%

confluency on 18 � 18 mm coverslips before fixation with

3.7% formaldehyde (Sigma) in phosphate buffered saline

(PBS). All washing steps after fixation were performed with

0.02% Tween in PBS (PBST). Cells were permeabilized with

0.5% Triton X-100 in PBS. GFP-fusion proteins were stained

with GFP-binding camelide-derived single chain antibody

fragments26 (GFP-binding protein; GBP) conjugated to

Cy3b, ATTO 550 or ATTO 647N (ChromoTek) diluted in

blocking solution (2% BSA/PBST). Alternatively, HeLa cells

were stained with primary mouse monoclonal antibodies

against Pan-histone (Roche) and secondary anti-mouse anti-

bodies labeled with Alexa 488, Alexa 594, Alexa 647 (Invitrogen).

Cells were counterstained with 100 ng ml�1 40,6-diamidino-2-

indole (DAPI) in PBST for 5 min and mounted on microscope

slides with indicated mounting media. In the case of the

ROXS-buffers oxygen was removed from the sample using

an oxygen-scavenging system based on glucose-oxidase/

catalase.27,28 The coverslip was sealed by mounting on a thin

strip of silicone paste applied on the microscope slide filled

with freshly prepared buffers or mounting media: (i) standard

ROXS-buffer with 1 mM concentrations of ascorbic acid (AA)

and 1 mM N,N0-methyl viologen (MV);16 (ii) ROXS-buffer

containing 1 mMTrolox (TX) that was additionally illuminated

with UV-light (lamp-filter) for 10 min in order to form Trolox-

quinone (TQ) in sufficient concentrations 425 mM;9,10 (iii)

pure Vectashield (VS) medium (Vector Labs);14 (iv) Ibidi-MM

mounting-medium (Ibidi).

Photobleaching was performed with an UltraVIEW VoX

spinning disc microscope equipped with an integrated FRAP

PhotoKinesis accessory (PerkinElmer), assembled to an Axio

Observer D1 inverted stand (Carl Zeiss) and using an oil

immersion objective (63�, NA 1.40, PlanApo, Carl Zeiss).

Confocal time series were recorded with intervals of 1 frame/s,

exposure times of 100–200 ms, frame size of 1024� 1024 pixels

and 14-bit image depth. For each experiment the entire field of

view containing approximately 5 to 20 cell nuclei was selected

for bleaching. Depending on the tested fluorophore either one

of the 488 nm, 561 nm or 635 nm laser line (measured laser

power at the objective: 4.5 mW, 2,1 mW and 1,1 mW,

respectively) was used for bleaching and imaging with

appropriately adjusted AOTF settings (typically 50 or 100%

for bleaching with the focused laser and 5–10% for imaging

through the spinning disk). Typically 10 prebleach frames were

followed by 30–50 repeated cycles of consecutive bleaching

and acquisition steps.

Quantitative evaluation was performed using ImageJ

(http://rsb.info.nih.gov/ij/). First, the nuclear area of cells

within the bleached region was selected either manually or

by threshold-based segmentation. Then the mean fluorescence

intensity over time was extracted, background subtracted and

normalized to the mean of the last five prebleach values. For

one experiment, 20–50 cells in three to five independent time

series were investigated per buffer condition. Values obtained

in different media conditions were normalized with respect to

intensity values in PBS. The mean values of fluorescence

intensity of one day of experiments were fitted with a bi-exponen-

tial decay function of the form y = A1 � exp(�x/t1) + A2 �exp(�x/t2) in Origin 8.0 (OriginLab). Experiments were

Dow

nloa

ded

by L

udw

ig M

axim

ilian

s U

nive

rsita

et M

uenc

hen

on 2

5/04

/201

3 13

:08:

05.

Publ

ishe

d on

11

Febr

uary

201

1 on

http

://pu

bs.r

sc.o

rg |

doi:1

0.10

39/C

0CP0

1919

D

View Article Online

Page 4: Citethis: Phys. Chem. Chem. Phys .,2011, PAPER€¦ · Citethis: Phys. Chem. Chem. Phys .,2011, ... aApplied Physics ... also find more specific effects that we summarize in practical

6702 Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 This journal is c the Owner Societies 2011

repeated at least twice on different days and mean values and

standard deviations of these experiments are given in the

supporting information. Values of photobleaching lifetime

were calculated according to tm = A1 � t1 + A2 � t2 with

respect to the value for PBS to correct for differences in

bleaching and imaging settings.

Sample and surface preparation for single-molecule

measurements

LabTek 8-well chambered cover slides (Nunc) with a volume

of B750 ml were treated with 0.1% HF for 30 s and were

washed three times with PBS. Subsequently, they were

incubated with a solution of 5 mg ml�1 BSA and 1 mg ml�1

BSA/biotin in PBS for at least 10 hours at 4 1C (all compounds

were used as received from Sigma Aldrich). After washing, the

surface was incubated with a B0.1 mg ml�1 solution of

streptavidin for 5 min and was washed three times with PBS.

Then a biotinylated 40mer oligonucleotide bearing Cy3B at

position 5 (30-CGT AT*A GCT ATA CAA TAT AAG TGT

AAG GAA TCG AAT CGT A-50 with T* = Cy3B (strand I);

as received from IBA, Germany) was incubated on the surface

for 1 min at high concentration (B10�8–10�9 mol l�1). Sub-

sequently, a complementary strand internally labeled with a

far-red dye at position 24 (50-GCA TAT CGA TAC ATT

AT*A TTC ACA TTC CTT AGC TTA GCA T-30 with

T* = ATTO 647N or ATTO 655 (strand II); as received

from IBA) was hybridized to strand I yielding a dsDNA probe.

Single-molecule experiments were carried out at room

temperature (22 � 1 1C): (i) If not stated otherwise standard

PBS pH 7.4 was used. (ii) If indicated, oxygen was removed

using an oxygen-scavenging system (PBS, pH 7.4, 10% (w/v)

glucose, 12.5% (v/v) glycerine, 50 mg ml�1 glucose-oxidase,

100–200 mg ml�1 catalase, and 0.1 mM Tris(2-carboxyethyl)-

phosphine hydrochloride). (iii) Different concentrations of

redox active agents or commercially available AFs were added

to the buffer as described in the text and figure captions.

Single-molecule fluorescence spectroscopy

A custom built confocal microscope was used as described in

ref. 19 to study single molecules. The excitation wavelength of

a pulsed supercontinuum-source (SuperK Extreme, Koheras)

was selected by a combination of an internal and an external

acousto-optical tunable filter (external AOTF: AOTFnc-VIS,

AA optoelectronic). The excitation wavelength for Cy3B was

533, and was set to 640 nm for ATTO 655 and ATTO 647N

(spectral width of 2 nm). After spatially filtering using a single-

mode fiber the beam was coupled into an oil immersion

objective (60�, NA 1.35, UPLSAPO 60XO, Olympus) of an

inverse microscope by a dual-band dichroic beam splitter for

recording fluorescent transients (Dualband z532/633, AHF

Analysentechnik). Average excitation intensities for all experi-

ments were 4.0 kW cm�2 at 533 nm and 1.5 kW cm�2 at

640 nm. The resulting fluorescence was collected by the same

objective, focused onto a 50 mm pinhole, filtered (Brightline

HC 582/75 for Cy3B, ET-Bandpass 700/75M for ATTO

655/ATTO 647N, both AHF Analysentechnik), and detected by

two avalanche photodiodes (SPCM-AQR-14, PerkinElmer).

Custom made LabVIEW software was used to register the

signal and for subsequent data evaluation.

The following procedure was conducted to obtain ON-counts

and OFF-times of single-molecule transients: (i) an auto-

correlation of the respective fluorescence transients was generated.

(ii) the autocorrelation-curve was fitted using an exponential

function; (iii) the OFF-times toff and ON-counts Non were

derived from the amplitudes and the characteristic time-

constant of the autocorrelation after background correction

according to ref. 19. On average B30 molecules were measured

and evaluated for each data point as described for each experi-

mental condition.

3. Results and discussion

The four different AFs that we compared in this study are

depicted in Fig. 2. Two of the AFs are ROXS based, one of

which contains the reductant ascorbic acid (AA) and the

oxidant N,N-methyl viologen (MV). The ubiquitous oxidant

oxygen is removed enzymatically. This buffer system is referred

to as ROXS (AA/MV). The second ROXS based AF consists

of the reductant Trolox (TX) and uses its quinone derivate

Trolox-quinone (TQ) as oxidant in combination with enzymatic

oxygen removal and is referred to as ROXS (TX/TQ). The

typical concentration of each component (AA/MV/TX) in the

respective buffer is 1 mM. TQ was shown to be formed in

concentrations of 425 mM after dissolving TX in aqueous

buffer or under UV-illumination and hence doesn’t have to be

added to the buffer.9,10,16

Both combinations of ROXS-buffers (AA/MV or TX/TQ)

could already prove their suitability to stabilize or alter the

fluorescence properties of single fluorophores by photo-induced

redox-reactions and are intensively used in single-molecule

spectroscopy and super-resolution imaging.9,10,16,29,30

The commercial AFs, namely Vectashield (VS) and Ibidi

mounting medium (Ibidi-MM) contain phenylenediamine

(PPD) and DABCO, respectively. In addition, these mounting

media contain between 70 and 90% glycerol. Thus, they are

highly viscous and intrinsically exhibit at least a 3-fold reduced

oxygen concentration compared to water.31 This means that

the ROXS buffers provide the additional advantage of a lower

oxygen concentration in the samples contributing to a more

efficient suppression of photobleaching. We believe, however,

that the reduced oxygen concentration in glycerol plus the

Fig. 2 Chemical structures of the main compounds found in com-

mercial AFs and ROXS-buffers: a) ROXS with the reductant AA and

oxidant MV, ROXS (AA/MV); b) ROXS with the reductant TX and

oxidant TQ, ROXS (TX/TQ); c) main compound of the commercial

AF Vectashield: reductant PPD; main compound of the commercial

Ibidi-MM: reductant DABCO.

Dow

nloa

ded

by L

udw

ig M

axim

ilian

s U

nive

rsita

et M

uenc

hen

on 2

5/04

/201

3 13

:08:

05.

Publ

ishe

d on

11

Febr

uary

201

1 on

http

://pu

bs.r

sc.o

rg |

doi:1

0.10

39/C

0CP0

1919

D

View Article Online

Page 5: Citethis: Phys. Chem. Chem. Phys .,2011, PAPER€¦ · Citethis: Phys. Chem. Chem. Phys .,2011, ... aApplied Physics ... also find more specific effects that we summarize in practical

This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 6703

reduced mobility lead to comparable conditions for ROXS

and commercial AFs as also in samples with enzymatic oxygen

removal the final oxygen concentration only reaches the lower

micromolar range.32

To compare the different AFs for immunofluorescence

we performed photobleaching experiments in a commercial

confocal microscope frequently used for biological studies

(UltraVIEW VoX, PerkinElmer, details see Materials and

Methods). Human HeLa cells grown on microscope coverslips

were immunolabeled with antibodies conjugated to different

fluorophores. Fluorophore-coupled antibodies against GFP

were used to label expressed GFP-tagged histone H2B-GFP

with an organic dye of choice (Cy3B, ATTO 550, ATTO

647N). Histone H2B-GFP is stably incorporated into nucleosomes

of chromatin and is evenly distributed within the nucleus.

Alternatively, histones were detected with primary mouse

monoclonal antibodies against Pan-histone and secondary

anti-mouse antibodies conjugated to the respective dyes

(Alexa 488, Alex 594, Alexa 647). Regions with cells were imaged

and subsequently bleached in iterative cycles; the average

fluorescence intensity of bleached regions was registered after

each cycle.

ATTO 647N as an example for carborhodamine dyes

Fig. 3 shows exemplary images of typical bleaching experi-

ments with the fluorophore ATTO 647N. This carborhodamine

type dye (structure see Scheme 1) is known for its outstanding

photophysical performance (emax = 150 000 l mol�1 cm�1,

ffl = 0.65 as stated on the ATTO-TEC homepage,

www.atto-tec.com) especially on the single-molecule level.33

Fluorescent nuclei of cells embedded in either PBS or

Vectashield show hardly any remaining fluorescence after

40 bleaching cycles reflecting ensemble bleaching of the fluoro-

phore (Fig. 3a/b). In contrast, when using ROXS (AA/MV) as

buffer medium only partial bleaching of ATTO 647N is

observed (Fig. 3c) indicating a stronger resistance against

photobleaching.

For quantitative evaluation the mean fluorescence intensity

over time was determined (details see Material and Methods).

On average for each buffer condition, 10 independent time

series (10–20 nuclei each) were recorded at least on two

different days (Fig. 4).

The reference bleaching behaviour of PBS is represented by

black squares in Fig. 4a; the intensity is reduced to E0.25

of its original value after E35 bleaching cycles (bc). The

experimental data is well reproduced by a bi-exponential decay

(solid lines in Fig. 4a). Results from repetitive experiments and

fitting of these data are summarized in Table S1; fit curves

from a representative experiment are shown in Fig. 4a together

with experimental data. According to these results, the

fluorophores bleach with two different bleaching com-

ponents and a mean bleaching lifetime tm = 23 � 2 bc with

tm = A1 � t1 + A2 � t2. A non-exponential behavior may be

attributed to differing microenvironments of the fluorophores

and has been reported before.15

VS shows a very similar relative bleaching lifetime of

1.3� 0.6 with respect to PBS but with a considerable amplitude

of a fast decay component and overall large experimental errors

between different measurements (Fig. 4a, circles; Table S1).y Asignificant increase of the photobleaching resistance of ATTO

647N is observed for Ibidi-MM, ROXS (AA/MV) and ROXS

(TX/TQ) with mean bleaching lifetimes over several experi-

ments of 2.3 � 0.2, 2.9 � 0.5 and 3.1 � 0.5 relative to PBS,

respectively (Table S1). In all cases the fraction of the fast

bleaching component is significantly reduced to E10%

(Fig. 4a). These data quantitatively evidence that both ROXS

buffers and Ibidi-MM are able to significantly enhance photo-

stability of ATTO 647N in fluorescence imaging.

In the next step we compare the mean brightness of

the fluorophores, i.e., of fresh samples (mounted shortly

before performing photobleaching experiments) and aged

samples (after 3 days storage at 4 1C) for the various media

Fig. 3 Confocal images of HeLa histone H2B-GFP expressing cells

labeled with ATTO 647N-coupled GFP antibodies. Samples were

embedded using different mounting media. Selected images of a time

series with (a) PBS (b) Vectashield and (c) ROXS (AA/MV) are shown

here exemplarily. The ATTO 647N dye was excited at 635 nm (close to

its absorption maximum at E644 nm). The panels show fluorescently

labelled nuclei before and after 40 iterative cycles of confocal imaging

and bleaching the entire field of view. Significantly less fluorescence

intensity is left in the PBS- and VS-embedded samples, whereas ROXS

effectively preserves fluorescence. Bar is 20 mm.

y Note that all relative bleaching constants in Table S1 and in the textare derived from several measurement days, while the data in theFig. 4a/5a/6a/S1a/S2/S3a) show representative mean bleaching curvesfrom a particular experimental day.

Dow

nloa

ded

by L

udw

ig M

axim

ilian

s U

nive

rsita

et M

uenc

hen

on 2

5/04

/201

3 13

:08:

05.

Publ

ishe

d on

11

Febr

uary

201

1 on

http

://pu

bs.r

sc.o

rg |

doi:1

0.10

39/C

0CP0

1919

D

View Article Online

Page 6: Citethis: Phys. Chem. Chem. Phys .,2011, PAPER€¦ · Citethis: Phys. Chem. Chem. Phys .,2011, ... aApplied Physics ... also find more specific effects that we summarize in practical

6704 Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 This journal is c the Owner Societies 2011

(Fig. 4b). For a systematic comparison we have chosen an

intermediate ‘‘storage-time’’ of 3 days to be able to detect

significant changes in fluorescence intensities that would allow

predictions on the general suitability of the respective medium

to preserve fluorescence upon storage. We are convinced that

the binding capabilities of antibodies do not play a significant

role in our case for the following reasons. (i) The samples were

4% formaldehyde fixed, a generally accepted way to stably fix

and preserve biological structures. (ii) We used approved,

highly affine antibodies (see Material and Methods) and we

did not observe any unbound fluorescent precipitates or a

suspicious increase in background fluorescence after long-term

storage that would hint to a dissociation of antibodies from

their epitopes. (iii) We know from long-standing experience

with similar immunofluorescently-stained samples that they

can be stored for many months when either mounted in

glycerol-based antifades or in PBS without obvious changes

in fluorescence intensities or structure.

While VS and Ibidi-MM reduce the average fluorescence

intensity for a freshly prepared solution to 70% and 40%

relative to PBS, respectively, both ROXS-AFs preserve the

brightness of ATTO 647N. A moderate decrease of ATTO

647N brightness was observed after storage in both ROXS

media, while storage in VS and PBS did not significantly affect

fluorescence intensity. Interestingly, a recovery of mean

brightness was observed in Ibidi-MM embedded samples,

reaching about the same level as VS.

These data provide valuable information for the fluoro-

phore choice for the desired imaging application considering

(i) photobleaching resistance of the fluorophore in the

AF medium (Fig. 4a) and quantified by the obtained fit-

parameters (see below), as well as (ii) fluorescence brightness

of the fluorophores in the AF medium and (iii) storage

stability of the fluorophores in the AF medium (Fig. 4b).

While the first parameters are obvious quality criteria in

fluorescence microscopy many applications might also require

storage stability of the sample. In particular for immuno-

fluorescence staining experiments, adherent cells, tissue or

whole mount samples are typically embedded in a layer of

mounting medium (either hardening or glycerol based and

sealed with nail polish) between a microscope slide and a

coverslip. Such samples should ideally be able to be stored for

days or weeks without major loss of quality.

Rhodamine dyes

To check for the general validity of the findings for ATTO 647N

we repeated the measurements for different fluorophores from

the dye classes of rhodamines and cyanines, with absorption/

emission spectra ranging from the blue to red spectral range.

For Alexa 488, a fluorophore with absorption/emission in

the blue/green region of the visible spectrum,23 results from

photobleaching experiments are depicted in Fig. 5a.

All experimental decays are again well-described by

bi-exponential functions; fit results are shown in Table S1.

Surprisingly, we do not observe significant improvement

of the photostability for both Ibidi-MM (Fig. 5a, stars;

relative tmean=1.3 � 0.2) and ROXS (AA/MV) buffer

(Fig. 5a, triangles; tmean=1.1 � 0.1) with respect to PBS.

While the ROXS (TX/TQ) buffer increases the relative tmean

by 1.8 � 0.2 (Fig. 5a, inverted triangles) we observe the best

performance for VS with a relative tmean = 2.9 � 0.4 with

respect to PBS (Fig. 5a, squares). The fluorescence brightness

of Alexa 488 is similar to PBS for all AFs (Fig. 5b). However,

storage in ROXS (AA/MV) leads to a substantial loss in

fluorescence (o50%), while in VS, Ibidi-MM and ROXS

(TX/TQ) mounted samples show no significant decrease of

brightness.

Other investigated rhodamine dyes were ATTO 550, a

fluorophore with absorption and emission in the green/yellow

spectral range, and Alexa 594, a fluorophore with excitation

and emission maxima in the orange/red spectral range. Both

show a similar trend in terms of photostabilizing abilities and

fluorophore brightness as reported in the previous sections: for

ATTO 550 all AFs significantly increase photobleaching

resistance compared to PBS (Fig. S1a, squares and fit; Table S1)

by 1.2 for Ibidi-MM, E2.5 for VS and ROXS (AA/MV) up to

a maximum of 3.2 for ROXS (TX/TQ); results are summarized

Fig. 4 Results from photobleaching experiments of the dye ATTO 647N obtained from confocal microscopy. (a) Time-course of the normalized

fluorescence intensity over 40 bleaching cycles for different AFs: PBS, squares; VS, circles; Ibidi-MM, stars; ROXS (AA/MV), triangles; ROXS

(TX/TQ), inverted triangles. Mean curves of at least five measurements from one particular measurement day are shown. Standard deviations were

typically below 0.02 and are thus omitted for clarity. Solid lines show a bi-exponential fit with obtained fit-parameters as listed in Table S1.

(b) Relative mean preableach fluorescence intensities for fresh samples and for aged samples mounted in different AFs (day 0 = fresh sample;

day 3 = sample aged for three days at 4 1C). Error bars indicate the standard error of mean intensities.

Dow

nloa

ded

by L

udw

ig M

axim

ilian

s U

nive

rsita

et M

uenc

hen

on 2

5/04

/201

3 13

:08:

05.

Publ

ishe

d on

11

Febr

uary

201

1 on

http

://pu

bs.r

sc.o

rg |

doi:1

0.10

39/C

0CP0

1919

D

View Article Online

Page 7: Citethis: Phys. Chem. Chem. Phys .,2011, PAPER€¦ · Citethis: Phys. Chem. Chem. Phys .,2011, ... aApplied Physics ... also find more specific effects that we summarize in practical

This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 6705

in Fig. S1a and Table S1. We note a significant enhancement

of the brightness of ATTO 550 in Ibidi-MM compared to

other AFs. Upon storage in both ROXS media we observe a

similar drop in fluorescence as for Alexa 488 (Fig S1b).

Experiments with Alexa 594 also follow these described trends:

all AFs have a positive influence on the photobleaching

resistance of the dye with moderate increases by a factor of

1.5 � 0.1 for ROXS (AA/MV) with respect to PBS (Fig. S2,

Table S1). Both VS and ROXS (TX/TQ) show a signifi-

cant increase by factors of 2.5 and 2.9, respectively, while

Ibidi-MM performs outstandingly well for Alexa 594 with

only 15% fluorescence loss compared to PBS (Fig. S2a, note

that the data could not be fitted to a bi-exponential function).

Again a moderate increase of brightness is observed for

Ibidi-MM and for ROXS (AA/MV) shortly after mounting

while the fluorescence intensity seems to be better preserved by

VS and Ibidi-MM when storing the sample (Fig. S2b).

Cyanine dyes

In the next section we investigated the influence of AFs on the

photophysical properties of cyanine dyes. As a first example

we chose Cy3B, a fluorophore with an absorption- and emission-

maximum at 558 nm and 572 nm, respectively. This dye is

frequently used in single-molecule FRET studies.34,35

As already found for rhodamine dyes all AFs clearly

increase the resistance of the fluorophore Cy3B against photo-

bleaching (Fig. 6a).

The mean bleaching lifetime of PBS (Fig. 6a, squares) is

increased by factors between 1.6 and 2.1 for ROXS (AA/MV),

Ibidi-MM, VS and ROXS (TX/TQ), compare Fig. 6a and

Table S1.

For all AF media we observed a significant decrease in

brightness of the fluorophore—a result that is surprising

considering that ROXS (AA/MV) was shown to increase the

brightness of surface-immobilized or diffusing fluorophores in

Fig. 5 Photobleaching experiments of the dye Alexa 488. (a) Time-course of the normalized fluorescence intensity over E25 bleaching cycles for

different buffers: PBS, squares; VS, circles; Ibidi-MM, stars; ROXS (AA/MV), triangles; ROXS (TX/TQ), inverted triangles. Mean curves of at

least five measurements from one particular measurement day are shown. Standard deviations were typically below 0.02 and are thus omitted for

clarity. Bi-exponential fits are shown as solid lines in the color of the respective data set. Fit results are summarized in Table S1. (b) Relative

fluorescence intensity for the different AFs before photobleaching and for aged samples (day 0 = fresh sample; day 3 = sample aged for three

days). Error bars indicate the standard error of mean intensities.

Fig. 6 Photobleaching experiments of the dye Cy3B. (a) The graph shows the time-course of the normalized fluorescence intensity over

E35 bleaching cycles for different AFs: PBS, squares; VS, circles; Ibidi-MM, stars; ROXS (AA/MV), triangles; ROXS (TX/TQ), inverted

triangles. Mean curves of at least five measurements from one particular measurement day are shown. Standard deviations were typically below

0.02 and are thus omitted for clarity. Bi-exponential fits are shown as solid lines in the color of the respective data set. Fit results are summarized in

Table S1. (b) Relative fluorescence intensity for the different AFs before photobleaching and for different aging-stages (day 0 = fresh sample;

day 3 = sample aged for three days). Error bars indicate the standard error of mean intensities.

Dow

nloa

ded

by L

udw

ig M

axim

ilian

s U

nive

rsita

et M

uenc

hen

on 2

5/04

/201

3 13

:08:

05.

Publ

ishe

d on

11

Febr

uary

201

1 on

http

://pu

bs.r

sc.o

rg |

doi:1

0.10

39/C

0CP0

1919

D

View Article Online

Page 8: Citethis: Phys. Chem. Chem. Phys .,2011, PAPER€¦ · Citethis: Phys. Chem. Chem. Phys .,2011, ... aApplied Physics ... also find more specific effects that we summarize in practical

6706 Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 This journal is c the Owner Societies 2011

single-molecule experiments.16 The brightness appeared to be

relatively stable after storage for all AFs (Fig. 6b).

Another commonly used cyanine dye is Alexa 647, a

dye that is very similar to Cy5 regarding its photophysical

properties as well as its molecular structure (see Scheme 1 and

ref. 25). This dye has an excitation maximum at 650 nm and

shows its emission maximum at 668 nm. Photobleaching

experiments reveal interesting findings (Fig. S3a, Table S1):

only moderate improvement of the photobleaching resistance

is observed for Ibidi-MM (factor of 1.1). The ROXS-buffers

improve both photobleaching resistance and fluorophore

brightness by factors of E3 (Fig. S3, Table S1). In contrast,

the fluorophore intensity was substantially reduced to less

than 20% in VS relative to PBS. This results in large experi-

mental errors for the determination of the photobleaching

time (data not shown). These facts indicate that this dye is not

well suited to be mounted with VS. The dye also shows poor

storage stability in all mounting media (Fig. S3b) and is hence

better substituted by ATTO 647N.

The presented data of dyes allow the following conclusions:

(i) The photostability of the investigated fluorophores

generally benefits from the addition of either of the AFs

(VS/Ibidi-MM/ROXS) independently from the spectral

properties of the dye. (ii) They undergo moderate to substantial

increase in their photobleaching resistance with the best

average AF-performance of ROXS (TX/TQ), which shows a

substantial increase in the photobleaching resistance for all

six studied dyes (Table S1). (iii) Fluorophore brightness is

typically unaffected or only slightly changed in all AFs com-

pared to PBS alone except for a major decrease observed for

Alexa 647 in VS. (iv) Both glycerol-based mounting media,

VS and Ibidi-MM, seem better suited for storage of the

studied fluorophores compared to ROXS media due to their

brightness conserving properties.

Single-molecule sensing of redox properties of commercial AFs

The quantitative comparison of AF characteristics in their

application using fluorescently labelled biological samples

revealed preferences for different dyes and quite specific inter-

actions such as the fluorescence loss of Alexa 647 when used

with VS. Whereas the ROXS AFs function is described by

a photophysical model, DABCO (main component in

Ibidi-MM) and PPD (main component in VS) were selected

empirically. To address the underlying photophysical and

physico-chemical mechanisms and to elucidate whether a

ROXS-like mechanism might also be involved with these

empirically found AFs, we carried out single-molecule

measurements with the SMRS ATTO 655. We therefore

recorded single-molecule transients of ATTO 655 labelled

DNA, immobilized via biotin/streptavidin on BSA surfaces

and analyzed the blinking kinetics of ATTO 655 at different

AF concentrations. It has been shown before,10,19 that the

number of photons emitted during one ON-period Non is an

inverse measure for the reducing properties of the environment

and that the OFF-times toff are a measure of the oxidizing

properties. Increasing concentrations of AA, for example, at

constant concentrations of the oxidant MV, reduce the number

of Non since the fluorophore is transferred to its radical anion

state more rapidly. Increasing concentrations of MV, on the

other hand, reduce toff.19

Fig. 7 shows representative fluorescence transients of

surface-immobilized ATTO 655 molecules in PBS (Fig. 7a)

and with different concentrations of VS-medium (Fig. 7b–e).

The fluorescence of ATTO 655 in PBS in the absence of redox-

active agents is comparatively stable with rare OFF-states

and no amplitude in the autocorrelation function down to 1

ms (not shown).Adding VS to the solution at a concentration of 0.07% v/v

results in the appearance of blinking with average values of

ton = 450 � 190 ms and toff = 990 � 560 ms (Fig. 7b shows

a representative transient; sensor values were derived from

on/off histograms). Since the excitation intensity was the same

for all measurements, ton is proportional to Non and is used in

our discussion. The OFF-state is attributed to a reduction

reaction into a metastable radical anion. Since oxygen that

also acts as an oxidant is not removed in these measurements

and because 0.07% VS does not alter the oxygen solubility and

diffusion constant significantly, the value of 990 ms has to be

compared to the literature value of toff = (1800� 1000) ms for

PBS with AA.19 Increasing the concentration of VS to 0.22%

v/v changes both values significantly to ton = 75 � 9 ms and

toff = 625 � 155 ms (Fig. 7c shows a representative trace;

average sensor values derived from on/off histograms).

Further increase to 0.70% v/v VS-stock solution changes

the sensor values to ton = 30 � 4 ms, toff = 390 � 150 ms

(Fig. 7d; sensor values derived from autocorrelation analysis

of intensity transients). The 15-fold reduction of ton reflects

the known reducing properties of PPD. The simultaneous

3-fold reduction of toff indicates the presence of some oxidizing

properties of the AF medium. The decrease of toff is certainly

Fig. 7 Fluorescence transients of the single-molecule redox sensor

ATTO 655 in PBS and at varying concentrations of VS (details see

Materials and Methods): (a) PBS, (b) PBS with 0.07% v/v VS-stock

solution (SMRS-values: ton = 450 � 190 ms, toff = 990 � 560 ms;

derived from on/off histograms), (c) PBS with 0.22% v/v VS-stock

solution (SMRS-values: ton = 75� 9 ms, toff = 625� 155 ms; derived

from on/off histograms), (d) PBS with 0.70% v/v VS-stock solution

(SMRS-values: ton = 30 � 4 ms, toff = 390 � 150 ms; derived from

autocorrelation analysis).

Dow

nloa

ded

by L

udw

ig M

axim

ilian

s U

nive

rsita

et M

uenc

hen

on 2

5/04

/201

3 13

:08:

05.

Publ

ishe

d on

11

Febr

uary

201

1 on

http

://pu

bs.r

sc.o

rg |

doi:1

0.10

39/C

0CP0

1919

D

View Article Online

Page 9: Citethis: Phys. Chem. Chem. Phys .,2011, PAPER€¦ · Citethis: Phys. Chem. Chem. Phys .,2011, ... aApplied Physics ... also find more specific effects that we summarize in practical

This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 6707

significant and might be weakened by reduced oxygen solubility

and mobility at higher VS fractions. The amount of oxidant in

the solution is substantially smaller than that of the reductant

and may hence be attributed to oxidizing impurities or chemical

products of PPD upon dissolving in buffer or glycerol.

These results and their interpretation are supported by

similar experiments with the dye ATTO 647N: single, immobilized

molecules show extremely stable fluorescence in ROXS

(AA/MV) as well as in ROXS (TX/TQ); Fig. 8a). The addition

of small amounts of VS to a deoxygenated PBS solution

(Fig. 8b) does not increase photostability in a similar manner;

here strong blinking of ATTO 647N is observed with short

ON-times and long OFF-periods (Fig. 8b) similar to observa-

tion in the absence of reductant due to triplet blinking or in the

presence of only reductants due to radical ion formation.

Replacing PBS by pure VS, however, already shows drastic

reduction of blinking and allows ATTO 647N to fluoresce for

a longer time-period (Fig. 8c).

These results are in accordance with the expectations of the

ROXS concept. In order to fluoresce stably, the dye ATTO

647N needs a sufficient amount of both oxidant and reductant.

For low concentration of VS the need for an oxidant is

not fulfilled as the concentration is too low, while at higher

concentration both redox-compounds are present in sufficient

concentrations.

Next, we carried out analogous measurements for Ibidi-MM

(Fig. 9). We also find a reduction of both sensor values upon

increasing concentration of Ibidi-MM pointing to similar

properties of this DABCO containing medium compared to VS.

Quantitatively, however, the properties of Ibidi-MM are

quite different to VS since ton only changes 3.7-fold whereas

toff is reduced 15-fold. These data indicate that Ibidi-MM has

substantial oxidizing properties and comparatively mild reducing

properties.

The results presented in the current section clearly show

that VS and Ibidi-MM might have been selected because in

the buffer applied they exhibit both reducing and oxidizing

properties and thus imply some aspects of the ROXS concept.

At this point we do not know the origin of the reducing and

oxidizing properties and can only speculate whether the forma-

tion of oxidizing species is related to partial degradation

of the original reducing compound. As mentioned above, such

a mechanism has been demonstrated for Trolox yielding

Trolox-quinone.10 Our investigations hence show that the

principle of triplet-state quenching by redox-active agents

and recovery of the fluorophore’s ionized states by subsequent

redox-reactions (the ROXS concept, see Fig. 1) is of fundamental

importance to prevent photobleaching in all fluorescence

applications. Our study also reveals that, though the chemical

and physical reasons for the successful application of several

AFs were unknown, their working action can be traced back

to a combination of oxidizing and reducing agents.

4. Conclusions: which Antifade for which

fluorophore?

In this paper we compared the performance of commonly used

and commercially available antifades such as Vectashield and

Ibidi-MM with new AFs based on the recently developed

ROXS concept. We employed an application-oriented assay

using confocal imaging and bleaching of fluorescently labeled

cells, and complemented it by single-molecule spectroscopy.

We found that generally all AFs exhibit a significant photo-

stabilizing effect. It turned out that no optimal AF could be

identified but that different AFs are preferential for different

fluorophores and different conditions. Our single-molecule

study revealed that this can at least in part be explained by

the fact that the commercial AFs Vectashield und Ibidi-MM

also exhibit reducing and oxidizing properties and thus use

principles of the ROXS concept. These results clearly demon-

strate the importance of the ROXS concept for fluorescence

applications in general and also show how the empirical

development of commercial antifading agents might have

lead to the use of compounds that have both reducing and

oxidizing capabilities.

The differences between the tested AFs might be related to

the different concentrations, mobility and reactivity of the

involved redox agents on the one hand as well as to more

Fig. 8 Fluorescence transients of single ATTO 647N molecules in

(a) ROXS buffer (1 mM AA/MV, enzymatic oxygen removal and

(b) 0.14% v/v VS-stock solution dissolved PBS with enzymatic oxygen

removal and (c) pure VS-stock solution.

Fig. 9 Fluorescence transients of ATTO 655 in PBS for varying

concentrations of Ibidi-MM: a) PBS with 0.1% v/v Ibidi-MM-stock

solution (SMRS-values: ton = 2200� 1100 ms, toff = 1100� 600 ms),

(b) PBS with 7% v/v Ibidi-MM-stock solution (SMRS-values:

ton = 1600 � 700 ms, toff = 710 � 200 ms), (c) PBS with 33%

v/v Ibidi-MM-stock solution (SMRS-values: ton = 600 � 170 ms,

toff= 75� 40ms). All SMRS values were derived from on/off-histograms

of single-molecules (E30–40).

Dow

nloa

ded

by L

udw

ig M

axim

ilian

s U

nive

rsita

et M

uenc

hen

on 2

5/04

/201

3 13

:08:

05.

Publ

ishe

d on

11

Febr

uary

201

1 on

http

://pu

bs.r

sc.o

rg |

doi:1

0.10

39/C

0CP0

1919

D

View Article Online

Page 10: Citethis: Phys. Chem. Chem. Phys .,2011, PAPER€¦ · Citethis: Phys. Chem. Chem. Phys .,2011, ... aApplied Physics ... also find more specific effects that we summarize in practical

6708 Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 This journal is c the Owner Societies 2011

specific interactions between certain AFs and fluorophores

that in some cases even lead to near complete quenching of a

fluorophore. A generalized overview with the aim of providing

practical guidance to microscopists also including further

observations that could not be discussed in detail in the

context of this work is provided in the following.

Vectashield performs best with the green emitting dye Alexa

488 (widely used for multicolour applications), performs well

with dyes in the orange-red emission range (Cy3b, ATTO 555,

Alexa 594) and is compatible with imaging of cyan emitting

dyes, such as the DNA binding DAPI and Hoechst (data not

shown). In contrast, the DABCO containing Ibidi-MM

performs significantly better with far-red emitting dyes and

similar or slightly worse than VS in combination with most

other dyes. Thus, it offers a good compromise in multicolour

application of improved photostability throughout the entire

spectrum. The relative high diffraction index (1.440) due to

their glycerol-based formulae allows deeper imaging into the

sample when using high NA oil objectives. Both commercial

AFs provide good storage stability for routine use with typical

biological samples.

The ROXS based formulae show comparable or slightly

better antifading performance with orange and red emitting

dyes and they clearly outperform both glycerol-based media

with far-red emitting dyes (Alexa 647 and ATTO 647N). On

the other hand, they show only little to moderate effect with

green emitting Alexa 488. This indicates that the multi-photon

processes become more and more influential with decreasing

excitation wavelength. For example in the case of Alexa

488 and ATTO 550, ROXS seems to work less effective com-

pared to red-absorbing dyes, a fact that points to an increasing

influence of higher-excited states that contribute to the

overall bleaching next to the triplet-route, which is effectively

suppressed by ROXS. This interpretation has also been

corroborated by other studies exciting, for example, red dyes

simultaneously with blue and red light.33,36,37

In the direct comparison of redox active agents, TX/TQ

appeared in most cases superior to AA/MV in terms of

antifading efficiency and storage stability. The antifading

performance of ROXS-AFs is, strongly dependent on the

oxygen-free environment in aqueous solution, which requires

extra sealing of the embedded sample, e.g., with silicone. Storage

may be compromised by evaporation and the performance is

less reproducible due to variable degrees of oxygen depletion.

The lower refractive index of the aqueous solution may reduce

image quality due to the refractive index mismatch with the

immersion oil. This should be considered in particular when

imaging deeper (410 mm) into the sample using high NA

oil objectives, whereas near-field applications such as TIRF

(often used in single molecule applications, e.g. single-molecule

tracking or single molecule localization microscopy) are not

compromised. For wide-field applications the usage of water-

immersion objectives may be advantageous with these AFs.

Recently, 2,20-thiodiethanol (TDE) embedding was reported

to be superior for stimulated emission depletion (STED)

microscopy of ATTO 647N labeled samples by having refractive

index of 1.52, equal to immersion oil.38 In our hands, however,

cells embedded in TDE show a flattened morphology similar

to the hardening media. While orange and red emitting dyes

(e.g., Alexa 555, ATTO 647N) show enhanced fluorescence,

other dyes in the lower spectral range (e.g., DAPI, Alexa 488)

show strongly decreased brightness in TDE (data not shown).

All tests in this study have been performed on organic dyes

conjugated to antibodies. It should be noted that all analyzed

AFs had no significant influence on photostability of green

fluorescent protein (GFP) when excited with the 488 nm

laser line (data not shown). This finding might be related

to the fact that the chromophore formed by the tripeptide

Ser65–Tyr66–Gly67 is located protected within the beta-barrel

structure of the protein.39

Bright and photostable fluorescence reduces exposure-time

and increases the signal-to-noise ratio, and is therefore an

important prerequisite for high quality imaging of biological

samples. This is even more true for applications of recent

super-resolution imaging methods, such as single-molecule

localization microscopy (e.g., STORM, PALM),7,33 STED6

and 3-dimensional structured illuminationmicroscopy (3D-SIM)40

that are generally more demanding for the employed dyes. The

choice of the right antifade medium makes an important

contribution to the dye performances. This paper gives useful

guidelines for their application and provides a framework for

further optimization of antifading formulae of mounting

media on the basis of the ROXS principle.

Acknowledgements

The authors thank C. Eggeling for discussions and helpful

information. We thank U. Rothbauer (ChromoTek GmbH)

for providing fluorophore coupled GFP antibodies. We are

grateful to S. Holzl for technical assistance. We are indebted to

H. Leonhardt for generous support. T. Cordes is supported

by a Marie-Curie Intra-European Fellowship provided by

the European Commission under the Seventh Framework

Programme (grant PIEF-GA-2009-255075). The project was

supported by the DFG (Inst 86/1051-1 to P. Tinnefeld and

SFB TR5 to L. Schermelleh), the Nanosystems Initiative

Munich (NIM) and the BioImaging Network Munich.

References

1 B. N. G. Giepmans, S. R. Adams, M. H. Ellisman and R. Y. Tsien,Science, 2006, 312, 217–224.

2 W. E. Moerner and L. Kador, Phys. Rev. Lett., 1989, 62,2535–2538.

3 M. Orrit and J. Bernard, Phys. Rev. Lett., 1990, 65, 2716–2719.4 E. Betzig, G. H. Patterson, R. Sougrat, O. W. Lindwasser,S. Olenych, J. S. Bonifacino, M. W. Davidson, J. Lippincott-Schwartz and H. F. Hess, Science, 2006, 313, 1642–1645.

5 M. J. Rust, M. Bates and X. Zhuang, Nat. Methods, 2006, 3,793–795.

6 S. W. Hell, Science, 2007, 316, 1153–1158.7 L. Schermelleh, R. Heintzmann and H. Leonhardt, J. Cell Biol.,2010, 190, 165–175.

8 M. Fernandez-Suarez and A. Y. Ting, Nat. Rev. Mol. Cell Biol.,2008, 9, 929–943.

9 I. Rasnik, S. A. McKinney and T. Ha, Nat. Methods, 2006, 3,891–893.

10 T. Cordes, J. Vogelsang and P. Tinnefeld, J. Am. Chem. Soc., 2009,131, 5018.

11 A. N. Fletcher and M. E. Pietrak, Appl. Phys. B: Photophys. LaserChem., 1985, 37, 151–157.

12 J. Widengren, A. Chmyrov, C. Eggeling, P. A. Lofdahl and C. A. M.Seidel, J. Phys. Chem. A, 2007, 111, 429–440.

Dow

nloa

ded

by L

udw

ig M

axim

ilian

s U

nive

rsita

et M

uenc

hen

on 2

5/04

/201

3 13

:08:

05.

Publ

ishe

d on

11

Febr

uary

201

1 on

http

://pu

bs.r

sc.o

rg |

doi:1

0.10

39/C

0CP0

1919

D

View Article Online

Page 11: Citethis: Phys. Chem. Chem. Phys .,2011, PAPER€¦ · Citethis: Phys. Chem. Chem. Phys .,2011, ... aApplied Physics ... also find more specific effects that we summarize in practical

This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 6699–6709 6709

13 R. Dave, D. S. Terry, J. B. Munro and S. C. Blanchard, Biophys.J., 2009, 96, 2371–2381.

14 A. Longin, C. Souchier, M. Ffrench and P. A. Bryon, Journal ofHistochemistry & Cytochemistry, 1993, 41, 1833–1840.

15 M. Ono, T. Murakami, A. Kudo, M. Isshiki, H. Sawada andA. Segawa, Journal of Histochemistry & Cytochemistry, 2001, 49,305–311.

16 J. Vogelsang, R. Kasper, C. Steinhauer, B. Person, M. Heilemann,M. Sauer and P. Tinnefeld, Angew. Chem., Int. Ed., 2008, 47,5465–5469.

17 T. Cordes, J. Vogelsang, M. Anaya, C. Spagnuolo, A. Gietl,W. Summerer, A. Herrmann, K. Mullen and P. Tinnefeld,J. Am. Chem. Soc., 2010, 132, 2404–2409.

18 T. Cordes, I. H. Stein, C. Forthmann, C. Steinhauer, M. Walz,W. Summerer, B. Person, J. Vogelsang and P. Tinnefeld, Proc.SPIE, 2009, 7367, 73671D.

19 J. Vogelsang, T. Cordes, C. Forthmann, C. Steinhauer andP. Tinnefeld, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 8107–8112.

20 C. Steinhauer, C. Forthmann, J. Vogelsang and P. Tinnefeld,J. Am. Chem. Soc., 2008, 130, 16840.

21 J. Vogelsang, T. Cordes, C. Forthmann, C. Steinhauer andP. Tinnefeld, Nano Lett., 2010, 10, 672–679.

22 C. Eggeling, C. Ringemann, R. Medda, G. Schwarzmann,K. Sandhoff, S. Polyakova, V. N. Belov, B. Hein, C. vonMiddendorff,A. Schonle and S. W. Hell, Nature, 2009, 457, 1159–U1121.

23 N. Panchuk-Voloshina, R. P. Haugland, J. Bishop-Stewart,M. K. Bhalgat, P. J. Millard, F. Mao, W. Y. Leung andR. P. Haugland, J. Histochem. Cytochem., 1999, 47, 1179–1188.

24 M. Cooper, A. Ebner, M. Briggs, M. Burrows, N. Gardner,R. Richardson and R. West, J. Fluoresc., 2004, 14, 145–150.

25 S. S. White, H. T. Li, R. J. Marsh, J. D. Piper, N. D. Leonczek,N. Nicolaou, A. J. Bain, L. M. Ying and D. Klenerman, J. Am.Chem. Soc., 2006, 128, 11423–11432.

26 U. Rothbauer, K. Zolghadr, S. Tillib, D. Nowak, L. Schermelleh,A. Gahl, N. Backmann, K. Conrath, S. Muyldermans,M. C. Cardoso and H. Leonhardt, Nat. Methods, 2006, 3, 887–889.

27 S. C. Blanchard, R. L. Gonzalez, H. D. Kim, S. Chu andJ. D. Puglisi, Nat. Struct. Mol. Biol., 2004, 11, 1008–1014.

28 S. C. Blanchard, H. D. Kim, R. L. Gonzalez, Jr., J. D. Puglisi andS. Chu, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 12893–12898.

29 A. Pertsinidis, Y. X. Zhang and S. Chu, Nature, 2010, 466,647–U611.

30 S. V. Solomatin, M. Greenfeld, S. Chu and D. Herschlag, Nature,2010, 463, 681–U117.

31 M. A. Joslyn and H. Supplee, Food Research, 1949, 14, 209–215.32 C. E. Aitken, R. A. Marshall and J. D. Pulglisi, Biophys. J., 2008,

94, 1826–1835.33 J. Vogelsang, C. Steinhauer, C. Forthmann, I. H. Stein,

B. Person-Skegro, T. Cordes and P. Tinnefeld, ChemPhysChem,2010, 11, 2475–2490.

34 J. Vogelsang, S. Doose, M. Sauer and P. Tinnefeld, Anal. Chem.,2007, 79, 7367–7375.

35 Y. Santoso, C. M. Joyce, O. Potapova, L. Le Reste, J. Hohlbein,J. P. Torella, N. D. F. Grindley and A. N. Kapanidis, Proc. Natl.Acad. Sci. U. S. A., 2010, 107, 715–720.

36 C. Eggeling, J. Widengren, L. Brand, J. Schaffer, S. Felekyan andC. A. Seidel, J. Phys. Chem. A, 2006, 110, 2979–2995.

37 X. X. Kong, E. Nir, K. Hamadani and S. Weiss, J. Am. Chem.Soc., 2007, 129, 4643–4654.

38 T. Staudt, M. C. Lang, R. Medda, J. Engelhardt and S. W. Hell,Microsc. Res. Tech., 2007, 70, 1–9.

39 R. Y. Tsien, Annu. Rev. Biochem., 1998, 67, 509–544.40 L. Schermelleh, P. M. Carlton, S. Haase, L. Shao, L. Winoto,

P. Kner, B. Burke, M. C. Cardoso, D. A. Agard,M. G. Gustafsson, H. Leonhardt and J. W. Sedat, Science, 2008,320, 1332–1336.

Dow

nloa

ded

by L

udw

ig M

axim

ilian

s U

nive

rsita

et M

uenc

hen

on 2

5/04

/201

3 13

:08:

05.

Publ

ishe

d on

11

Febr

uary

201

1 on

http

://pu

bs.r

sc.o

rg |

doi:1

0.10

39/C

0CP0

1919

D

View Article Online